| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Retired Forums
Forum: UD Windows Agent Support [Read Only]
Thread: Processor Speed Oddity |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 14
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi ooDAV,
   2) SiSoft Sandra at http://www.sisoftware.net/?location=update This download contains a number of free system diagnostic modules, including several standard benchmark routines. 3) Everest at http://www.lavalys.hu/products/overview.php?pid=1&lang=en This download is a freeware hardware diagnostics and memory benchmarking program that marysduby posted about. Try some more benchmarks, on clean boots without WCG running, etc. Maybe they will prove that the UD benchmark is a  Or maybe they will confirm astounding facts about your motherboards.   mycrofth |
|||
|
||||
|
RT
Master Cruncher USA - Texas - DFW Joined: Dec 22, 2004 Post Count: 2636 Status: Offline Project Badges: 
|
Sounds to me like the P4 has a problem of some sort. As you correctly point out, it should outperform the Celeron. As Rick indicated, the benchmark is just a bunch of instructions that are run and timed. So that should not be an issue.
----------------------------------------All said and done, the Celeron is performing better during the benchmarks than the P4. That along with the fact that the P4 is underperforming among its P4 peers tells me the P4 has a problem of some sort. Either it is the Mobo as someone suggested, the workload running during the benchmarks (could be a Trojan, virus, firewall or some other task) or it could be something kicking in the throttling. So, the solution to the problem is likely in one of those three areas. The suggestions of mycrofth sound good to me. Still I don't understand why, if the celeron is running the benchmark faster, how the P4 is running the WU faster. Things are not adding up here. Wait a minute... You are not making the mistake that I have many times... you are not looking at "overall" score are you? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Dear ... colleagues, I just want to help you understand me.
----------------------------------------Let me tell you a bit more. I'm IT professional. Worked as programmer (3 years) and system administrator (7 years) I've assembled hundred of PCs and installed OS/applications on hundreds of PCs. I know words "Trojan", "virus" and "throttling" not only from forums and blogs. 63°C on Prescott is not enough for throttling. Nevertheless, I've used www.panopsys.com/downloads/ThrottleWatch.zip Yes. I'm talking about processor performance score, not overall. I've switched processors between computers and got the same result for P4 on another motherboard, so you can't tell me that motherboard or memory or whatever else has a bottleneck. SiSoft Sandra was on my test list. (And yes, agents was stopped :) ) here is Sandra results: Celeron 2.8: SiSoftware Sandra Benchmark Results CPU Arithmetic Benchmark: Dhrystone ALU : 7637 MIPS Whetstone iSSE2 : 3711 MFLOPS CPU Multimedia Benchmark: Integer x8 iSSE2 : 17314 it/s Float x4 iSSE2 : 21781 it/s Results Interpretation: Higher index values are better. <skip> Processor Model : Intel(R) Celeron(R) CPU 2.80GHz Speed : 2.80GHz Performance Rating : PR3081 (estimated) Cores per Processor : 1 Unit(s) L2 On-board Cache : 128kB ECC Synchronous, ATC, 2-way set, 64 byte line size, 2 lines per sector Chipset 1 Model : ASUSTeK Computer Inc 82865G/PE/P, 82848P DRAM Controller / Host-Hub Interface Front Side Bus Speed : 4x 100MHz (400MHz data rate) Features (W)MMX Technology : Yes SSE Technology : Yes SSE2 Technology : Yes SSE3 Technology : No SSE4 Technology : No HTT - Hyper-Threading Technology : No P4 Prescott 3.0 SiSoftware Sandra Benchmark Results CPU Arithmetic Benchmark: Dhrystone ALU : 8744 MIPS Whetstone iSSE3 : 6309 MFLOPS CPU Multimedia Benchmark Integer x8 iSSE2 : 21289 it/s Float x4 iSSE2 : 28408 it/s <skip> Processor Model : Intel(R) Pentium(R) 4 CPU 3.00GHz Speed : 3.01GHz Performance Rating : PR4005 (estimated) Cores per Processor : 1 Unit(s) L2 On-board Cache : 1MB ECC Synchronous, ATC, 8-way set, 64 byte line size, 2 lines per sector, 2 threads sharing Chipset 1 Model : Intel Corporation 82915G/GV/GL/P/PL/GL Grantsdale Host Bridge/DRAM Controller Front Side Bus Speed : 4x 201MHz (804MHz data rate) Features SSE2 Technology : Yes SSE3 Technology : Yes SSE4 Technology : No HTT - Hyper-Threading Technology : Yes btw, after 4h13m Celeron: 42% Prescott: 71% It can be a bug in Windows Processor Driver (yes, processor has a driver now :) ) Or UD benchmark code has different routines for different processor family/stepping/etc. Anyway it's now a fact that UD agent benchmark has a problem with Prescott on my machine. And benchmark result has nothing to do with work unit processing speed. I'll do tests on similar Prescott machines. I'll do more tests with BOINC for windows on the same machines (it's pity that it's not fully supported agent for now) I whant to ask you to do simillar tests also. Use P4 Prescott and compare it to NorthWood Yours sincerely /Andrey [Edit 1 times, last edit by Former Member at Dec 16, 2005 2:42:12 PM] |
||
|
|
RT
Master Cruncher USA - Texas - DFW Joined: Dec 22, 2004 Post Count: 2636 Status: Offline Project Badges:
|
Does anyone know if the benchmark requires a larger data array than is normal in the WUs - hence causing the problems if the on-chip cache is crippled for some strange reason?
---------------------------------------- |
||
|
|
|